We have recently found that primary melanomas on the palms and soles arise from a field of histologically normal but genetically abnormal melanocytes surrounding the tumors. The genetic abnormalities in the 'field cells'are present in the in situ and invasive portions of the tumor. Thus the field cells are a subtle precursor of melanoma in situ, and are a possible source of local recurrence in melanoma since they are not detectable by current methods, and may extend outside the empirically-determined excision margins used in clinical practice (Bastian et al., 2000a). The evidence for this in acral melanomas is very strong. Improved understanding of the mechanism of the local recurrence in other types of melanoma would have significant clinical impact. Lentigo maligna melanoma (LMM) is a form of melanoma it shares with acral melanoma its histological growth pattern and risk of recurrence. We have found evidence for the existence of field cells in LMM, and hypothesize that they can cause recurrences when present at the margin. We will test this hypothesis by identifying DNA copy number changes in the invasive components of LMM using array CGH and look for their presence in the normal appearing melanocytes surrounding these tumors using FISH. We will determine whether the presence of field cells at the surgical margins is predictive of local recurrence. A critical question relates to the initial genetic events that permit the clonal growth of abnormal melanocytes out of which the tumor eventually develops. LMM arises on the chronically sun-exposed skin of elderly individuals. LMM incidence is significantly increased in individuals with xeroderma pigmentosum (XP). Our preliminary data indicates frequent deletions in genomic regions harboring DNA repair genes in LMM. We hypothesize that in LMM one of the earliest events causes a defect in the response to Uv-induced DNA. We will analyze genomic DNA of LMMs for recurrent aberrations of genomic regions harboring UV response genes. Candidate genes within these regions will undergo a mutational analysis for the detection of acquired or inherited sequence variations. UV response genes found to be altered in LMMs will be functionally evaluated in vitro and in vivo using xenografted artificial human skin reconstructs exposed to UV light. The long-term goals of this project is the understanding of the genetic progression of LMM and the identification of markers that improve the detection of minimal residual melanoma and current recommendations for safety margins.